Usefulness of a Metal Artifact Reduction Algorithm in Digital Tomosynthesis Using a Combination of Hybrid Generative Adversarial Networks

In this study, a novel combination of hybrid generative adversarial networks (GANs) comprising cycle-consistent GAN, pix2pix, and (mask pyramid network) MPN (CGpM-metal artifact reduction [MAR]), was developed using projection data to reduce metal artifacts and the radiation dose during digital tomo...

Full description

Bibliographic Details
Main Authors: Tsutomu Gomi, Rina Sakai, Hidetake Hara, Yusuke Watanabe, Shinya Mizukami
Format: Article
Language:English
Published: MDPI AG 2021-09-01
Series:Diagnostics
Subjects:
Online Access:https://www.mdpi.com/2075-4418/11/9/1629
id doaj-e4ae731aa7264c7697e0bef0382a6d33
record_format Article
spelling doaj-e4ae731aa7264c7697e0bef0382a6d332021-09-25T23:59:12ZengMDPI AGDiagnostics2075-44182021-09-01111629162910.3390/diagnostics11091629Usefulness of a Metal Artifact Reduction Algorithm in Digital Tomosynthesis Using a Combination of Hybrid Generative Adversarial NetworksTsutomu Gomi0Rina Sakai1Hidetake Hara2Yusuke Watanabe3Shinya Mizukami4School of Allied Health Sciences, Kitasato University, Sagamihara 252-0373, Kanagawa, JapanSchool of Allied Health Sciences, Kitasato University, Sagamihara 252-0373, Kanagawa, JapanSchool of Allied Health Sciences, Kitasato University, Sagamihara 252-0373, Kanagawa, JapanSchool of Allied Health Sciences, Kitasato University, Sagamihara 252-0373, Kanagawa, JapanSchool of Allied Health Sciences, Kitasato University, Sagamihara 252-0373, Kanagawa, JapanIn this study, a novel combination of hybrid generative adversarial networks (GANs) comprising cycle-consistent GAN, pix2pix, and (mask pyramid network) MPN (CGpM-metal artifact reduction [MAR]), was developed using projection data to reduce metal artifacts and the radiation dose during digital tomosynthesis. The CGpM-MAR algorithm was compared with the conventional filtered back projection (FBP) without MAR, FBP with MAR, and convolutional neural network MAR. The MAR rates were compared using the artifact index (AI) and Gumbel distribution of the largest variation analysis using a prosthesis phantom at various radiation doses. The novel CGpM-MAR yielded an adequately effective overall performance in terms of AI. The resulting images yielded good results independently of the type of metal used in the prosthesis phantom (<i>p</i> < 0.05) and good artifact removal at 55% radiation-dose reduction. Furthermore, the CGpM-MAR represented the minimum in the model with the largest variation at 55% radiation-dose reduction. Regarding the AI and Gumbel distribution analysis, the novel CGpM-MAR yielded superior MAR when compared with the conventional reconstruction algorithms with and without MAR at 55% radiation-dose reduction and presented features most similar to the reference FBP. CGpM-MAR presents a promising method for metal artifact and radiation-dose reduction in clinical practice.https://www.mdpi.com/2075-4418/11/9/1629tomosynthesismetal artifact reductiongenerative adversarial networkarthroplastyradiation-dose reduction
collection DOAJ
language English
format Article
sources DOAJ
author Tsutomu Gomi
Rina Sakai
Hidetake Hara
Yusuke Watanabe
Shinya Mizukami
spellingShingle Tsutomu Gomi
Rina Sakai
Hidetake Hara
Yusuke Watanabe
Shinya Mizukami
Usefulness of a Metal Artifact Reduction Algorithm in Digital Tomosynthesis Using a Combination of Hybrid Generative Adversarial Networks
Diagnostics
tomosynthesis
metal artifact reduction
generative adversarial network
arthroplasty
radiation-dose reduction
author_facet Tsutomu Gomi
Rina Sakai
Hidetake Hara
Yusuke Watanabe
Shinya Mizukami
author_sort Tsutomu Gomi
title Usefulness of a Metal Artifact Reduction Algorithm in Digital Tomosynthesis Using a Combination of Hybrid Generative Adversarial Networks
title_short Usefulness of a Metal Artifact Reduction Algorithm in Digital Tomosynthesis Using a Combination of Hybrid Generative Adversarial Networks
title_full Usefulness of a Metal Artifact Reduction Algorithm in Digital Tomosynthesis Using a Combination of Hybrid Generative Adversarial Networks
title_fullStr Usefulness of a Metal Artifact Reduction Algorithm in Digital Tomosynthesis Using a Combination of Hybrid Generative Adversarial Networks
title_full_unstemmed Usefulness of a Metal Artifact Reduction Algorithm in Digital Tomosynthesis Using a Combination of Hybrid Generative Adversarial Networks
title_sort usefulness of a metal artifact reduction algorithm in digital tomosynthesis using a combination of hybrid generative adversarial networks
publisher MDPI AG
series Diagnostics
issn 2075-4418
publishDate 2021-09-01
description In this study, a novel combination of hybrid generative adversarial networks (GANs) comprising cycle-consistent GAN, pix2pix, and (mask pyramid network) MPN (CGpM-metal artifact reduction [MAR]), was developed using projection data to reduce metal artifacts and the radiation dose during digital tomosynthesis. The CGpM-MAR algorithm was compared with the conventional filtered back projection (FBP) without MAR, FBP with MAR, and convolutional neural network MAR. The MAR rates were compared using the artifact index (AI) and Gumbel distribution of the largest variation analysis using a prosthesis phantom at various radiation doses. The novel CGpM-MAR yielded an adequately effective overall performance in terms of AI. The resulting images yielded good results independently of the type of metal used in the prosthesis phantom (<i>p</i> < 0.05) and good artifact removal at 55% radiation-dose reduction. Furthermore, the CGpM-MAR represented the minimum in the model with the largest variation at 55% radiation-dose reduction. Regarding the AI and Gumbel distribution analysis, the novel CGpM-MAR yielded superior MAR when compared with the conventional reconstruction algorithms with and without MAR at 55% radiation-dose reduction and presented features most similar to the reference FBP. CGpM-MAR presents a promising method for metal artifact and radiation-dose reduction in clinical practice.
topic tomosynthesis
metal artifact reduction
generative adversarial network
arthroplasty
radiation-dose reduction
url https://www.mdpi.com/2075-4418/11/9/1629
work_keys_str_mv AT tsutomugomi usefulnessofametalartifactreductionalgorithmindigitaltomosynthesisusingacombinationofhybridgenerativeadversarialnetworks
AT rinasakai usefulnessofametalartifactreductionalgorithmindigitaltomosynthesisusingacombinationofhybridgenerativeadversarialnetworks
AT hidetakehara usefulnessofametalartifactreductionalgorithmindigitaltomosynthesisusingacombinationofhybridgenerativeadversarialnetworks
AT yusukewatanabe usefulnessofametalartifactreductionalgorithmindigitaltomosynthesisusingacombinationofhybridgenerativeadversarialnetworks
AT shinyamizukami usefulnessofametalartifactreductionalgorithmindigitaltomosynthesisusingacombinationofhybridgenerativeadversarialnetworks
_version_ 1717367425914109952